Presentation at the Annual Meeting of the Linguistic Association of the Southwest

We presented some of our work at the Linguistics Association of the Southwest annual meeting which was held virtually in September 2020. You can find our presentation below. In addition, we answer some of the questions that we received during our talk, in an effort to make our work more accessible. If you have any further questions, please don’t hesitate to contact us!

Cite as: Campos-Astorkiza, R., O. Muxika-Loitzate, K. DeLeon, K. LoCascio and S. Sullivan 2020. How early should we teach pronunciation? Sound category formation in beginner and intermediate learners. Presented at the Annual Meeting of the Linguistics Association of the Southwest, Sept 24-26, 2010

LASSO 2020_presentation

  • Have you considered measuring other acoustic cues, given the fact that native speakers of Spanish might present voicing of voiceless stops?

So far, we have not looked at other acoustic cues. However, we have tried to capture any variation in the type of production for the voiceless stops by categorizing each token in our study according to the type of realization, including voiceless stops, voiced stop, tap, approximant, frication and deletion. This categorical analysis showed that the great majority of productions were voiceless stops. There were very few taps and approximants. This results are explained in the presentation. Moving forward we might need to consider whether using other acoustic cues might be necessary given the lack of variation in the type of production. However, we might consider other cues that might distinguish voiceless stops in English and Spanish such as burst intensity or the effect on the duration of the preceding vowel.

  • Do you think your results would be different if you treated proficiency level as a continuous variable?

Proficiency level is treated a binary variable in our study: beginner vs. intermediate. This categorization is based on the college course that our participants were taking, either a third-semester language course or an upper level Spanish pronunciation class. This means that we didn’t measure proficiency level but rather used course as a proxy for proficiency. This was based in the expectation that the level of Spanish in these two classes in very different and we can safely assume that two groups of students present very different abilities in Spanish.

If proficiency level was treated as a continuous variable we might expect to find the same direction for the effect, i.e. as the level becomes higher, there might be a lesser decrease in VOT. However, we do acknowledge that this direction of effect might revert at some point in a leaner’s acquisition path since they might be able to move beyond the learning plateau and decrease their VOT durations even more than beginners.

  • Did you have a control group that didn’t receive any pronunciation instruction?

We did not have a control group. The reason is that the research question is answered by comparing the effect of pronunciation in two different groups of participants. The question is not whether pronunciation affects voiceless stops production but whether it has a different degree of effect depending on the proficiency of the learners.